CD19 antibody-drug conjugate therapy in DLBCL does not preclude subsequent responses to CD19-directed CAR T-cell therapy

TO THE EDITOR:

Diffuse large B-cell lymphoma (DLBCL) is the most common histologic subtype of non-Hodgkin lymphoma (NHL), with aggressive clinical behavior.¹  Approximately 40% to 50% of DLBCL patients are refractory to or relapse after frontline chemoimmunotherapies.²  Management of relapsed/refractory DLBCL is challenging, and treatment options include salvage therapy followed by autologous hematopoietic cell transplantation (autoHCT), in younger, chemosensitive patients.³  CD19-directed chimeric antigen receptor (CAR) modified T-cell therapy is another potentially curative option for patients with relapse after autologous hematopoietic cell transplantation or those with refractory disease.⁴  Axicabtagene ciloleucel and tisagenlecleucel are 2 anti-CD19 CAR T-cell treatments approved by the US Food and Drug Administration for relapsed/refractory DLBCL.^5,6 

The CD19 antigen is an attractive target for immunotherapy in B-cell non-Hodgkin lymphoma. It is expressed during B-cell development only after B-lineage commitment and is thus not present on hematopoietic stem cells.⁷  CD19 expression is lost during terminal plasma cell differentiation but maintained in hematologic B-cell malignancies.⁷  Several clinical trials using monoclonal antibodies,⁸  antibody-drug conjugates (ADCs),⁹  and bispecific T-cell engagers targeting the CD19 antigen are ongoing in B-cell non-Hodgkin lymphoma. Studies investigating CD19-directed CAR T-cell therapies in aggressive lymphomas frequently excluded patients previously treated with CD19 targeting immunotherapies. Hence, the feasibility and efficacy of anti-CD19 CARs in lymphoma patients with prior CD19-directed immunotherapies are not known. Loncastuximab tesirine is an ADC comprising a humanized anti-CD19 monoclonal antibody stochastically conjugated to a pyrrolobenzodiazepine dimer toxin, SG3199.¹⁰  A phase 1 first-in-human study of loncastuximab tesirine demonstrated encouraging clinical activity in patients with relapsed/refractory DLBCL,⁹  and a phase 2 study recently finished patient accrual (#NCT03589469).

We sought to evaluate the outcomes of anti-CD19 CAR T-cell therapy in relapsed, refractory DLBCL previously treated with CD19-directed immunotherapy. Adult (age ≥18 years) DLBCL patients were identified from 2 multicenter, open-label studies of loncastuximab tesirine (phase 1: #NCT02669017 and phase 2: #NCT03589469), who subsequently received anti-CD19 CAR T-cell therapy. This retrospective analysis was approved by the institutional review board. Deidentified patient data were collected in collaboration with 6 academic medical centers (US centers = 4; United Kingdom = 1; Italy = 1) involved in the loncastuximab tesirine trials. Cytokine release syndrome was graded using Lee et al 2014 criteria,¹¹  and neurotoxicity was graded as per Common Terminology Criteria for Adverse Events, version 5.0.

A total of 14 DLBCL patients with disease relapsing or progressing after treatment with loncastuximab tesirine and subsequently undergoing CD19-directed CAR T-cell therapy were identified (Table 1). Among the 14 patients, 11 patients (79%) were male, and 13 patients (93%) were white. The median age was 58.5 years (range, 27 to 86). Ten had de novo DLBCL (germinal center B-cell–like = 4; non–germinal center B-cell = 2; not known = 4), and 4 patients had DLBCL transforming from indolent histologies (marginal zone lymphoma = 1; follicular lymphoma = 1; nodular lymphocyte predominant Hodgkin lymphoma = 1). Five patients (36%) had a high-intermediate international prognostic index at the time of diagnosis. c-MYC gene rearrangement was identified in 3 patients (21%) (1 patient had triple-hit lymphoma), whereas c-MYC status was unknown in 3 patients (21%). The median interval between diagnosis of DLBCL and initiation of loncastuximab tesirine was 21.5 months (range, 6.8 to 258). These patients received a median of 2 cycles (range, 1 to 7) of loncastuximab tesirine. The antitumor response to loncastuximab tesirine in these 14 patients was as follows: 8 patients (57%) had refractory disease, 5 patients (36%) attained partial response, and 1 patient (7%) achieved a complete response (overall response rate [ORR] = 43%). All responding patients had progression of the disease before proceeding with CAR therapy.

Table 2 summarizes details of CAR T-cell therapy in these subjects. The median interval between loncastuximab tesirine and CAR T-cell therapy was 120 days (range, 22 to 600). Six patients received additional lines of therapy between loncastuximab tesirine and CAR T-cell treatment (median of 1 therapy line; range, 1 to 3). The CD19 expression was assessed by immunohistochemical staining on repeat biopsies in 10 patients (71%) in between loncastuximab tesirine and CAR administration. All 10 tested patients were positive for CD19 after ADC failure. This information was not available in 4 patients (29%). Before CAR T-cell administration, 13 patients had refractory or progressive disease, whereas 1 patient was in partial remission. All patients received standard lymphodepletion with fludarabine and cyclophosphamide before CAR T-cell therapy. The type of anti-CD19 CAR T-cell therapy received by the patients included axicabtagene ciloleucel (n = 5), tisagenlecleucel (n = 2), JCAR017 (n = 3), and investigational CARs targeting CD19 (n = 4; including 2 patients with CARs targeting dual antigens; Table 2). The median follow-up of survivors was 6 months (range, 3 to 22). Grade 1 to 2 cytokine release syndrome was common (n = 7; 50%). Grade 1 immune effector cell–associated neurotoxicity syndrome (ICANS) was identified in 4 patients (29%), and only 1 patient had grade 4 ICANS. Following CAR therapy, the best response at 3 months included 6 patients (43%) with a complete response and 1 patient (7%) with a partial response (ORR = 50%). Seven patients had refractory disease following CAR therapy. Five of 6 complete remissions are ongoing at a median of 6 months (range, 6 to 11). One patient with complete remission relapsed after 11 months and was alive at last follow-up (+22 months), whereas the patient achieving a partial remission subsequently died because of progressive lymphoma. Six out the 7 patients not achieving a CR expired at a median of 5 months (range, 1 to 9) after CAR therapy. None of the 2 patients receiving dual-antigen targeting CARs (targeting CD19/CD22 and CD19/CD20) achieved a response. All 4 patients with unknown CD19 expression status after loncastuximab tesirine achieved a complete remission with anti-CD19 CAR T-cell therapy, making it unlikely that these patients had relapsed with a CD19-negative disease.

This is the first report, to our knowledge, evaluating the efficacy of anti-CD19 CAR T-cell therapy after anti-CD19 immunotherapy. In this series of 14 cases, favorable outcomes of CAR T-cell therapy (ORR = 50%) were seen in patients with relapsed/refractory DLBCL after anti-CD19–targeted treatment with a pyrrolobenzodiazepine dimer–based ADCs. The toxicity profile of CAR treatment in these subjects appears consistent with published data.^5,6  CD19 expression on DLBCL was reassessed in 10 patients after relapse or progression on loncastuximab tesirine treatment; no cases of CD19 antigen-negative relapse were seen. We wish to acknowledge that this report is limited by small sample size and retrospective design. In conclusion, our report suggests that prior treatment with anti-CD19 ADCs in relapsed/refractory DLBCL does not preclude subsequent responses to anti-CD19 CAR T-cell therapies. Additional data on the feasibility of CAR therapy in patients receiving CD19 monoclonal antibodies and bispecific T-cell engagers are needed.